EEG-based Emotion Recognition Using Recurrence Plot Analysis and K Nearest Neighbor Classifier
Publish place: 20th Iranian Conference on Biomedical Engineering(ICBME2013)
Publish Year: 1392
نوع سند: مقاله کنفرانسی
زبان: English
View: 905
This Paper With 6 Page And PDF Format Ready To Download
- Certificate
- من نویسنده این مقاله هستم
استخراج به نرم افزارهای پژوهشی:
شناسه ملی سند علمی:
ICBME20_093
تاریخ نمایه سازی: 25 فروردین 1394
Abstract:
Electroencephalogram (EEG)-based emotion recognition has been a rapidly growing field. However, accurate and sufficient performance rates are yet to be obtained. This paper presents the classification of EEG correlates on emotion using the relatively new non-linear feature extraction method, namely, Recurrence Plot analysis to extract thirteen non-linear features. This method is compared with feature extraction method based on spectral power analysis. The K nearest neighbor is applied to classify extracted features into the emotional states based on arousal-valence (high/low arousal, valence) plane with the addition of liking axis (positive/negative). Leading to performance rates of 58.05%, 64.56% and 67.42% for 3 classes of valence, arousal and liking; which confirm the advantage of a non-linear feature extraction method over previous frequency based feature extraction techniques
Keywords:
Authors
Fatemeh Bahari
Department of Biomedical Engineering Amirkabir University of Technology Tehran, Iran
Amin Janghorbani
Department of Biomedical Engineering Amirkabir University of Technology Tehran, Iran